Human resources departments manage one of the most crucial roles within a company, curating flows of candidates for new hires and shaping environments that help employees to succeed. The technology and data tools at HR offices has never been more advanced, but there are challenges ahead, both to adoption and using technology in responsible ways.
Ahead of our June “How Silicon Valley does HR” event, Orange Silicon Valley Human Resources Lead Wilson Lau sat down to field some high-level questions about what technology startups have to offer HR departments right now, as well as what startups and clients confront as they deploy new services using data science and artificial intelligence.
Orange Silicon Valley: What trends are driving change right now in the HR industry?
Wilson Lau: I think one of the big things happening in the HR world is people analytics. It’s really driving the landscape and change in HR operations. Particularly, we see a lot of companies — HR organizations — with new developments in data analytics and expertise where they build their own teams use different methodologies to collect data, to analyze data, and to build models to solve HR problems. And we see a lot of startups using the same approach to build a lot of new HR solutions that we’ve never seen before.
OSV: So from the data gathering, to the data storage, to what gets done with that data after it’s acquired — it’s fair to say those are all parts of what they’re working on.
WL: Exactly.
OSV: What kinds of benefits to new startups need to deliver to show their value to HR departments?
WL: For one company we’ve seen, they’re able to use their solution to solve a skill gap problem. In many cases, companies do their long-term planning and decide which goal they want to go after; that basically dictates all of the resources they need. So, how many people they need, what kind of skill set they need to acquire. The company can park information into their solution and get a skill gap analysis, which basically tells you what’s missing from the current workforce.
OSV: It’s an inventory, basically, of what a company has and what it needs.
WL: Right. So they can develop a training or recruitment program in order to fill those gaps.
OSV: Do the startups entering this space — or the HR departments considering this technology encounter any specific obstacles in terms of new technology adoption?
WL: Yes, one of the issues is potential bias built into algorithms. A lot of times the errors can be black boxes and the client cannot explain how the model works, knowing how it’s predicting one thing or another. And it’s becoming an issue in HR use cases. Just imagine that if you’re a recruiter you’re using this algorithm to automatically rank the candidates and select the candidates to go to the interview stage. If you have bias in the algorithm, the algorithm may potentially ignore a candidate who is a woman or minority — so you really have a problem with the HR organization if you cannot explain why this candidate is being selected or not being selected.
OSV: I see. If you’re on the client side being responsible for the decisions the algorithm is making, you still have to be able to speak to those and justify them.
WL: Right. Because otherwise you might get into some discrimination lawsuit or things like that.
OSV: Is there anything with data privacy — and perhaps GDPR — that’s becoming an issue?
WL: Definitely. In the case of GDPR, we know that it’s going into effect, and it’s regulating not just employee data, but also the contractors’ information and even candidates who submit resumes. So it’s putting a lot of pressure on the IT to upgrade its infrastructure in order to meet those requirements.
OSV: Are there any solutions that startups are exploring to deal with those issues?
WL: Yes, there are some startups already trying to develop solutions. Some of them are trying to address the bias issue that I mentioned earlier. Essentially, when we’re dealing with algorithms using some of the newer techniques like deep learning or neural network algorithms, some of the solutions that we know are able to provide insight knowing which neuron is being activated when a decision is made. So with that insight, their scientists could potentially do something about the algorithm and know better what is going on.
On GDPR, there are also some startups trying to make things easier by helping companies to manage consensus from the users — and also to manage employee data more quickly. That’s something that can be very helpful.
Do you like what you just read? Follow our events and learn about the latest trends in Silicon Valley from our tech and business analysts every week by subscribing to The Main Cable, Orange Silicon Valley’s email newsletter.